1. Field of the Invention
This invention relates generally to an electron switch for use in a communications network and, more particularly, to a free space electron switch for switching purposes in a communications network, where the switch employs an array of cathodes, each cathode being responsive to an optical or RF input signal on a communications channel and generating free space electrons in response thereto, and where the electron beams are selectively steerable by an aiming anode towards a particular receiver in an array of receivers associated with a plurality of output channels.
2. Discussion of the Related Art
State of the art telecommunications systems and networks typically employ optical fibers to transmit optical signals separated into optical packets carrying information over great distances. An optical fiber is an optical waveguide including a core having one index of refraction surrounded by a cladding having, another, lower, index of refraction so that optical signals propagating through the core at certain angles of incidence are trapped therein. Typical optical fibers are made of high purity silica including certain dopant atoms that control the index of refraction of the core and cladding.
The optical signals may be separated into channels to distinguish groups of information. Different techniques are known in the art to identify the channels through an optical fiber. These techniques include time-division multiplexing (TDM) and wavelength-division multiplexing (WDM). In TDM, different slots of time are allocated for the various packets of information. In WDM, different wavelengths of light are allocated for different data channels carrying the channels. More particularly, sub-bands of light within a certain bandwidth of light are separated by predetermined wavelengths to identify the various data channels.
When optical signals are transmitted over great distances through optical fibers, attenuation within the fibers reduces the optical signal strength. Therefore, detection of the optical signals over background noise becomes more difficult at the receiver. In order to overcome this problem, optical amplifiers are positioned at predetermined intervals along the fiber, for example, every 80-100 km, to provide optical signal gain. Various types of amplifiers are known in the art that provide an amplified replica of the optical signal, and provide amplification for the various modulation schemes and bit rates that are used.
Further, switching networks are periodically provided along the optical fiber path so that the optical packets of information can be switched and routed in a desired manner to reach their ultimate destination. Address bits within the optical packets provide location codes so that the switching devices can direct the optical packets to the appropriate optical fiber. Alternatively, the switching devices may be controlled by data that is provided out-of-band, via separate lines of communications. The switching networks at a particular location within the communications system may be required to support hundreds of thousands of data channels.
Because photons are highly non-reactive to the propagation medium of the optical fiber and to each other, providing suitable photon switching devices to redirect the optical signal is typically difficult. Further, pure optical switching is difficult to achieve because photons cannot be directed or steered without modifying the physical medium through which they propagate. Photon steering is typically accomplished by reflecting the photons off of moveable mirrors, or by passing the photons through LCD molecules or temperature-sensitive crystals. Because the process of modifying the physical medium to steer an optical beam tends to be slow and unwieldy, few photon switching technologies provide a fast enough response time necessary for state-of-the-art optical switching speeds, and those that may be fast enough typically cannot be scaled suitably to provide a sufficient number of output ports. Scalability defines the number of output ports that can be provided in the switch.
Moveable mirror switches employing micro-electrical mechanical systems (MEMS) are one type of mirror known in the art to switch an optical signal. Movable mirror switches of this type are typically separated into two categories, particularly, infinitely adjustable mirrors employing analog MEM switches, and two position mirrors employing digital MEMS switches. Digital MEMS switches potentially provide a relatively low switching speed (latency), but are not scaleable. Further, the number of internal components in a digital MEMS switch increases exponentially as the number of output ports increases, making them difficult to scale beyond a few hundred ports. Thus, large-scale MEMS switches typically employ adjustable analog mirrors that allow for greater scalability. However, analog MEMS switches typically have a high switching latency (low switching speed) requiring milliseconds to switch.
The longevity and reliability of MEMS switches for optical signal switching applications are suspect. Currently, a typical MEMS switch has a life on the order of one billion switching cycles. Therefore, if an analog MEM switch could operate fast enough to switch optical packets at commercially acceptable speeds, the switch would barely survive one minute before reaching the end of its operating life. Further, MEMS switches are sensitive to shocks, are fragile, and are bulky.
Additionally, current generation MEMS switches require the use of regenerator lasers, even in course, fiber-by-fiber switching applications, because of the lack of reflectivity of the mirrors. It is known to increase the reflectivity of the mirrors by, for example, gold plating the mirrors. However, it is not clear that this will eliminate the need for regenerator lasers, especially in real-world networks that have multiple hops and long-transmission lengths.
Practical lambda-by-lambda switching requires more than passively redirecting wavelengths from fiber to fiber. In order to prevent wavelength collisions, it is necessary to change the wavelength of the lambda switches as they hop from switch to switch. This requires the use of regenerator lasers. Tunable lasers do not mitigate this problem, because they still require that a given wavelength signal be reserved from end-to-end of the network. Employing wasting circuits, where the number of circuits is significantly increased beyond what is necessary to handle the required bandwidth, can possibly solve the collision problem.
Other photon switching technologies are available other than MEMS switches. These switching technologies include the Agilant bubble switch, LCD switches, switches that steer light using temperature-sensitive crystals, as well as other switches known in the art. However, all of these technologies typically suffer from lack of scalability and have a high switching latency.
Electronic switching offers an alternative to pure photonic switching. One well known technique of electronic switching for telecommunications systems employing optical fibers is by using single-stage crossbars. A crossbar is a semiconductor-based logic device that performs the switching operation. However, the number of internal components in a crossbar increases exponentially, or nearly exponentially, as the number of output ports increases. As a result, most crossbars have a maximum of 64 output ports. Next generation crossbars will provide a complex internal interconnect scheme that may allow the output port count to exceed 512 ports.
Crossbars are limited by the clock speed of their semiconductor logic gates, which is typically at or below 1 GHz. To obtain higher port speeds, multiple slower ports must be combined in order to create a single fast port, which greatly decreases the overall port count. For example, for a crossbar that runs at 622 MHz, 66 ports must be combined to create a single OC-768 port. Also, the de-multiplexers and multiplexers that separate the bit stream and then recombine the stream is complex and requires exotic switching technology, especially for OC-192 bit rates and beyond.
CLOS is an interconnection topology that allows smaller crossbars to be combined to form a larger, higher port count switch. Almost all existing and planned crossbar-based switches will be built using the CLOS topology. CLOS requires a large number of crossbars in order to obtain a given port count. As a result, CLOS interconnected crossbar switches have very large footprints, and consume a large amount of power.
Switching speed has also presented a problem with CLOS-based switches. Semiconductor CLOS switches typically require tens to hundreds of microseconds to establish a connection from an input port to an output port. Moreover, their switching speed is non-deterministic, in that the amount of time needed to establish a connection is highly unpredictable. For packet-by-packet switching applications, the complexity of the packet forwarding engines and traffic managers that control the switch is greatly increased because it is difficult for the switch to guarantee FIFO packet behavior. Further, unwanted effects are introduced into the output packet stream, such as jitter. It is likely that many of the CLOS crossbar-based electronic switches that are used within OEO optical cross-connects have such a slow switching speed and are highly unpredictable that they may not be suitable for packet-by-packet switching.
Board-to-board connector density is also a serious concern with CLOS switches. For large CLOS switches, nearly four out of every five interconnect is internal to the switch and cannot be used for external, through-switch bandwidth. Therefore, CLOS based switches are limited by the connector density, trace density and interconnects needed to create all of the internal intra-switch bandwidth. As a result, it has been suggested that CLOS switches hit hard limits in terms of board-to-board connector density at 512 output ports.
Also, because CLOS switches have a high component count, their reliability is an important concern. Any switch that consists of a full bay of ICs will support complex failure-recovery and re-routing capabilities. Also, as with all semiconductor logic-based switches, bit rate per port is limited by the clock rate of the logic gates.
Regardless of the drawbacks discussed above, the use of electrons are well suited for switching applications because electrons can be easily steered by electrostatic and electromagnetic fields. However, known electron switches have directed the electrons through digital logic gates within semiconductors. Thus, as discussed above, these devices have proved complex and difficult to scale for switching applications, and they are limited by the speed at which their solid-state logic gates are capable of switching.
In accordance with the teachings of the present invention, a communications system is disclosed that includes one or more free space electron switches. The free space electron switch employs an array of electron emitters, where each emitter is responsive to an RF or optical input signal on an input channel. Each emitter includes a cathode that emits electrons in response to the input signal. Each emitter further includes a focussing/accelerating electrode for collecting and accelerating the emitted electrons into an electron beam. Each emitter further includes an aiming anode that directs the beam of electrons to a desired detector within an array of detectors. The detector then converts the beam of electrons to a representative RF or optical signal on an output channel. The anodes are selectively steerable so that the electron beam is directed to the detector associated with a desired output channel to provide the switching.
The cathode can be a cold cathode, hot cathode or photocathode depending on the particular application. In one embodiment, the cathode is a photocathode that converts an optical input signal to an electron beam, and the detectors are e-beam lasers that convert electrons to an optical signal. In another embodiment, the cathode is a cold cathode that receives an electrical signal and generates the electron beam therefrom. In that embodiment, the emitter includes a modulating electrode that generates an electric field to modulate data onto the electron beam. Beam blanking can also be used to provide electron beam modulation.
The communications system employing the switch can be an ISDN, DSLAM networks, packet routing systems, ADSL networks, PBX systems, local exchange systems, etc. The free space electron switch also has application for use in a multiplexer or demultiplexer in the system.
Additional objects, advantages and features of the present invention will become apparent from the following description and appended claims, taken in conjunction with the accompanying drawings.